Delivering Stems
Splitting a mix into agreed groups of audio for mastering, film or live use, with formats and names that others can actually work with.
Related
Stems are a few agreed groups of audio that can be adjusted independently and still add up to something close to the mix. They are not multitracks, and not any bus you happen to export.
The problem
A mastering engineer wants to fix only the drums; a film mixer must duck music under dialogue; a game needs layers that switch by scene. One stereo mix is not enough, and dozens of multitracks are too much.
How it works
Stems are usually split by instrument group, such as drums, bass, guitars, keys, vocals and effects. Played together they should resemble the mix, so processing must be divided deliberately. Files start at the same time, run the same length, share sample rate and bit depth, and carry clear names.
Settings and judgement
A common delivery is 48 kHz and 24 bit. Leave peak headroom and do not limit individual stems. Include a note on bus processing, tempo and key. Keep an unprocessed set separate from the printed set.
What to listen for
Solo each stem, then play all together and check that the sum matches the mix. If one group jumps forward in the full set, the bus processing was not divided well.
Common misunderstandings
Stems are not multitracks. Multitracks are individual instruments; stems are groups. Stems should not be mastered separately either, because that breaks the balance when they are summed.
Listen in the library
All tracks from midicn-lib — inline badge is the license tier (C1 commercial / C2 non-commercial / C3 study only).
Sources
- Stems 与分轨、总线的区别，依影视声音与母带交付通识核查，未抄文本
- 48 kHz、24 bit 与等长起点等交付惯例，依通行后期交付规范核查，未抄文本
Used for fact-checking only; all prose is original.